1. Field of the Invention
The invention relates generally to the structure and fabrication process of semiconductor devices. More particularly, this invention relates to the structure and fabrication process of semiconductor devices that utilize shallow p-type regions.
2. Description of the Prior Art
As the overall dimensions of semiconductor devices are miniaturized and made ever smaller, the formation of very shallow p-doped regions, e.g., those less than a quarter-micron in depth, is becoming a major limiting factor in the fabrication process used to make metal oxide semiconductor field-effect transistors (MOSFET) and complementary metal oxide semiconductor (CMOS) devices.
The method used to make these vital CMOS and MOSFET transistors involves the formation of both n-type and p-type doped regions. Shallow n-doped regions can easily be formed by the ion implantation of arsenic or other n-type dopants. However, technical difficulties currently hamper the formation of very shallow p-doped regions. In most semiconductor fabrication processes the dopant boron is used to form p-type regions. Because boron has a very low atomic number (Z=5), a very low implantation energy must be used keep the dopant in the shallow surface region necessary for small geometry devices. Traditional ion implantation techniques are limited in their ability to produce low Z dopants with very low implantation energies. These limitations effect both the stability and reliability of shallow p-junctions. Furthermore, due to its low atomic number, implanted boron tends to channel through crystalline substrates and therefore forms a very undesirable, deep, `implantation profile tail.sup.1 ` in which the concentration of the dopant can not be controlled. This `channeling tail` creates great difficulties in the ability to clearly define the junction depth of boron implanted shallow p-type regions. This inability to control the junction depth seriously degrades device performance. All in all, it is not feasible to use conventional low energy implantation technology to implant boron and form shallow p-regions for very-large scale integrated (VLSI) circuits.
Several techniques are currently used to try to reduce the technical difficulties associated with low energy boron implantation. In one case the shallow p-region is formed using a heavier compound ion, i.e., BF2. Due to the higher mass of the compound, the constituent atoms of the BF2 ion have a shallower penetration depth for a given ion energy, thus enabling the formation of shallower p-type regions. The BF2 ions provide another key advantage because they help reduce problems caused by channeling effects. This improvement is accomplished by the increased crystal damage caused by the heavier fluorine component of the compound ion. However, since fluorine is neither a p-type nor n-type dopant, the fluorine atoms that are introduced from the BF2 ions do not directly contribute to the electrical performance of the semiconductor device.
The introduction of fluorine from the BF2 compound generates a new set of problems. Due to their low solubility in silicon, the fluorine atoms tend to migrate, particularly if the substrate is heated. After a BF2 implantation, any subsequent fabrication process which uses elevated temperatures, will tend to cause the implanted fluorine to migrate to the silicon surface, i.e., silicon-oxide interface. In some cases, this migration may cause the fluorine to coalesce and form a gap at the interface. Several different contact problems can be caused by migrating fluorine including; poor contact reliability, high contact resistance, and unstable electrical performance.
In addition to the contact problems caused the migration of fluorine, implantation of BF2 ions with energies less than 15 KeV can be difficult. This difficulty in turn limits the minimum depth of implantation and thus limits the miniaturization of integrated circuit (IC) devices in VLSI and ULSI applications. There are other difficulties associated with the implantation of BF2 ions including enhanced Si-epitaxial growth at the interface between silicon, i.e., the p+diffusion region, and an aluminum alloy containing silicon (a common interconnect material).
For all the above reasons, BF2 ion implantation is not a viable solution for the difficulties currently associated with the fabrication of shallow p-type regions. Therefore, there is a profound need in the art of IC device fabrication, particularly for devices requiring shallow p-type regions, to provide a fabrication process that will resolve these difficulties and limitations.